THE SPuUbpyY OF CANCER 15 
rising to 90 per cent. at the third transference. When once the 
initial difficulties have been overcome, the growth exhibited by 
propagated cancer, as illustrated by several of the tumours we 
have studied, soon passes beyond the bounds of all measure- 
ment. A small piece of tissue weighing no more than o’or grm. 
frequently increases to 1°5 grm. of tissue in ten days. This is 
an increase of over a hundredfold, and under suitable conditions 
go to 100 per cent. of the animals develop tumours through 
many successive transplantations. This sequence of events is 
not constant when tumours can be transplanted successfully. 
Some sporadic growths grow well from the outset of 
propagation. For others a marked increase in the success 
of transplantation is observed, frequently after three or four 
transferences. For others again the percentage of successful 
inoculations cannot be raised to, or maintained at, 100 per cent. ; 
but remains constant within certain fluctuations. However, for 
the majority of our carcinomata growth ceased at the third or 
fourth transference. We have interpreted those differences to 
signify primary differences in degree, inherent in the cells of 
different tumours. They persist during continued propagation 
in the majority of cases; so that each tumour exhibits from the 
outset of propagation a constant behaviour within limits. We 
have refrained from terming the increased success of trans- 
plantation occurring in special cases “an increase of virulence,” 
as has been done by analogy with what occurs when bacteria 
are propagated by the method of passage from one animal 
to another. The term “virulence” should be dropped and 
‘adaptability ” substituted, because the method of propagation 
is essentially an artificial selection of the most rapidly growing 
tumours, and their subdivision and distribution over a large 
number of animals. Although the greater success sometimes 
attending later inoculations suggests an alteration in the energy 
and rate of growth of the tumour cells, the phenomenon referred 
to may be explained without having recourse to this assump- 
tion: (1) by an increase in the number of proliferating cells, 
similar to what occurs in the propagation of domesticated 
animals as in the following example ; and (2) by adaptation. 
The increase in the numbers of any form of fancy dog 
which happens to become fashionable is due not to an artificial 
shortening of the period of canine gestation, but to increased 
breeding facilities, e.g. to multiplication of the number of those 
